Identification of Defective Maize Seeds Using Hyperspectral Imaging Combined with Deep Learning

被引:15
作者
Xu, Peng [1 ]
Sun, Wenbin [1 ]
Xu, Kang [1 ]
Zhang, Yunpeng [2 ]
Tan, Qian [2 ]
Qing, Yiren [2 ]
Yang, Ranbing [2 ]
机构
[1] Hainan Univ, Coll Informat & Commun Engn, Haikou 570228, Peoples R China
[2] Hainan Univ, Coll Mech & Elect Engn, Haikou 570228, Peoples R China
关键词
hyperspectral imaging; maize seeds; defect detection; feature selection; convolutional neural network; NONDESTRUCTIVE IDENTIFICATION; CLASSIFICATION; SELECTION;
D O I
10.3390/foods12010144
中图分类号
TS2 [食品工业];
学科分类号
0832 ;
摘要
Seed quality affects crop yield and the quality of agricultural products, and traditional identification methods are time-consuming, complex, and irreversibly destructive. This study aims to establish a fast, non-destructive, and effective approach for defect detection in maize seeds based on hyperspectral imaging (HSI) technology combined with deep learning. Raw spectra collected from maize seeds (200 each healthy and worm-eaten) were pre-processed using detrending (DE) and multiple scattering correction (MSC) to highlight the spectral differences between samples. A convolutional neural network architecture (CNN-FES) based on a feature selection mechanism was proposed according to the importance of wavelength in the target classification task. The results show that the subset of 24 feature wavelengths selected by the proposed CNN-FES can capture important feature information in the spectral data more effectively than the conventional successive projections algorithm (SPA) and competitive adaptive reweighted sampling (CARS) algorithms. In addition, a convolutional neural network architecture (CNN-ATM) based on an attentional classification mechanism was designed for one-dimensional spectral data classification and compared with three commonly used machine learning methods, linear discriminant analysis (LDA), random forest (RF), and support vector machine (SVM). The results show that the classification performance of the designed CNN-ATM on the full wavelength does not differ much from the above three methods, and the classification accuracy is above 90% on both the training and test sets. Meanwhile, the accuracy, sensitivity, and specificity of CNN-ATM based on feature wavelength modeling can reach up to 97.50%, 98.28%, and 96.77% at the highest, respectively. The study shows that hyperspectral imaging-based defect detection of maize seed is feasible and effective, and the proposed method has great potential for the processing and analysis of complex hyperspectral data.
引用
收藏
页数:21
相关论文
共 52 条
  • [1] Hyperspectral imaging coupled with multivariate analysis and artificial intelligence to the classification of maize kernels
    Alimohammadi, Fariba
    Rasekh, Mansour
    Sayyah, Amir Hosein Afkari
    Abbaspour-Gilandeh, Yousef
    Karami, Hamed
    Sharabiani, Vali Rasooli
    Fioravanti, Ambra
    Gancarz, Marek
    Findura, Pavol
    Kwasniewski, Dariusz
    [J]. INTERNATIONAL AGROPHYSICS, 2022, 36 (02) : 83 - 91
  • [2] Permutation importance: a corrected feature importance measure
    Altmann, Andre
    Tolosi, Laura
    Sander, Oliver
    Lengauer, Thomas
    [J]. BIOINFORMATICS, 2010, 26 (10) : 1340 - 1347
  • [3] The successive projections algorithm for variable selection in spectroscopic multicomponent analysis
    Araújo, MCU
    Saldanha, TCB
    Galvao, RKH
    Yoneyama, T
    Chame, HC
    Visani, V
    [J]. CHEMOMETRICS AND INTELLIGENT LABORATORY SYSTEMS, 2001, 57 (02) : 65 - 73
  • [4] Prediction of Sweet Corn Seed Germination Based on Hyperspectral Image Technology and Multivariate Data Regression
    Cui, Huawei
    Cheng, Zhishang
    Li, Peng
    Miao, Aimin
    [J]. SENSORS, 2020, 20 (17) : 1 - 11
  • [5] Recent Applications of Multispectral Imaging in Seed Phenotyping and Quality MonitoringAn Overview
    ElMasry, Gamal
    Mandour, Nasser
    Al-Rejaie, Salim
    Belin, Etienne
    Rousseau, David
    [J]. SENSORS, 2019, 19 (05)
  • [6] Hyperspectral imaging techniques for rapid detection of nutrient content of hydroponically grown lettuce cultivars
    Eshkabilov, Sulaymon
    Lee, Arim
    Sun, Xin
    Lee, Chiwon W.
    Simsek, Halis
    [J]. COMPUTERS AND ELECTRONICS IN AGRICULTURE, 2021, 181
  • [7] Hyperspectral imaging for seed quality and safety inspection: a review
    Feng, Lei
    Zhu, Susu
    Liu, Fei
    He, Yong
    Bao, Yidan
    Zhang, Chu
    [J]. PLANT METHODS, 2019, 15 (01)
  • [8] Hyperspectral Monitoring of Powdery Mildew Disease Severity in Wheat Based on Machine Learning
    Feng, Zi-Heng
    Wang, Lu-Yuan
    Yang, Zhe-Qing
    Zhang, Yan-Yan
    Li, Xiao
    Song, Li
    He, Li
    Duan, Jian-Zhao
    Feng, Wei
    [J]. FRONTIERS IN PLANT SCIENCE, 2022, 13
  • [9] Aflatoxin rapid detection based on hyperspectral with 1D-convolution neural network in the pixel level
    Gao, Jiyue
    Zhao, Longgang
    Li, Juan
    Deng, Limiao
    Ni, Jiangong
    Han, Zhongzhi
    [J]. FOOD CHEMISTRY, 2021, 360 (360)
  • [10] HyperSeed: An End-to-End Method to Process Hyperspectral Images of Seeds
    Gao, Tian
    Chandran, Anil Kumar Nalini
    Paul, Puneet
    Walia, Harkamal
    Yu, Hongfeng
    [J]. SENSORS, 2021, 21 (24)